Reed Hedges wrote:
> Well one of the easiest things to immediately grasp is simple key=value
> pairs.
And then they're always the good old INI format that everyone knows and
loves or hates, though it doesn't really do hierarchy so well
[objectname]
key = value
key = value
key = value
[objectna
Reed Hedges wrote:
> Peter Amstutz wrote:
>> It should be somewhat context-sensitive.
> avatar {
> types = misc:avatar
> misc:nick {
> # Since a property key is given (the loading code will have
> # a list of these, added by metaobjects), we can
> # assume this o
Peter Amstutz wrote:
> Well, I haven't had a chance to sit down and work on this, but basically
> I have a few principals I want to follow:
>
> 1) It should be very easy to write by hand. Requiring less typing than
> XML for example.
> 2) It should be readable and scan well, minimize extraneous
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Well, I haven't had a chance to sit down and work on this, but basically I
have a few principals I want to follow:
1) It should be very easy to write by hand. Requiring less typing than
XML for example.
2) It should be readable and scan well, m
Peter Amstutz wrote:
> On Tue, 18 Oct 2005, Lalo Martins wrote:
>
>>> may I suggest YAML (www.yaml.org) or a subset thereof?
>
>
> It's a thought. Despite support for a bunch of scripting languages,
> it's not clear if there is a C library available for it. However I
> can't say I'm thrilled b
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Tue, 18 Oct 2005, Lalo Martins wrote:
may I suggest YAML (www.yaml.org) or a subset thereof?
It's a thought. Despite support for a bunch of scripting languages, it's
not clear if there is a C library available for it. However I can't say
I'
